Relationship between esophageal and alveolar pressure variations-during occlusion.

To the Editor: Failure of mouth pressure variations (APm) to reflect alveolar pressure variations (APA) during thoracic gas volume (TGV) measurements by Boyle’s law has been reported in asthmatic patients by Shore and Martin (1) and quite recently by Stanescu et al. (2). In both studies the conclusions were reached by comparing TGV’s obtained by relating plethysmographic volume changes (AVbox) to APm (TGVm = AVbox/APm) and by relating them to pressure variations in the esophagus (TGVes = AVbox/APes), taken to represent pleural pressure variations (APpl). How close APpl is to APA is the question. It has been discussed by Stanescu et al. (2) on the basis of a monoalveolar model featuring lung and airways properties, including the compliance of intrathoracic airway walls and the shunt impedance of the upper airways. This model study showed that for normal compliances of lung tissue (CL), alveolar gas (Cg), intrathoracic airway wall (CB), and extrathoracic airway wall (CE), APpl was only 3% larger than APA at a frequency of 2 Hz. Moreover APP~/APA, in contrast with APm/APA, was not substantially modified by increasing peripheral and central airway resistance and by decreasing the impedance of extrathoracic airway walls. Having made similar computer simulation both with the same and with a two-compartment model, I would like to add a few points. Monoalveolar model. In static conditions, i.e., at infinitely low panting frequency, APpl and APA are in phase and AP~~/APA = 1 + (Cg + CE)/(CL + CB). It may be seen that APpl may be substantially larger than APA when Cg is large and CL is small (when specific lung compliance is low, e.g., at high lung volume in normal subjects). AP~~/APA is also directly related to upper airway wall compliance (CE). Although the relationship between APpl and APA is given by a high-order differential equation, their amplitude ratio and phase angle (+) may be predicted to vary little with the panting frequency. In all the situations examined by Stanescu et al. (2) and up to 4 Hz, APpl/ APA does not differ by more than 1% from its static value, as given above, and I+] is less than 1". Bialveolar model. I assumed that the lung was made of two compartments arranged in parallel, each with its tissue compliance (Cl, C,), gas compliance (Cgl, Cgz), and airway resistance (RI, Rz). Tissue resistances and airway walls compliances were neglected. In such a model APA will not be the same in the two compartments (APA~ # APA~) except in static conditions or when their specific compliances are equal (CJCg, = CJCgz). The accuracy of TGVes will depend on the ability of APpl to reflect the weighted average of local APA [APA = (APA~. Cg, + APA~. Cg2)/(Cg, + Cgz)]. The relationship between APpl and APA is given by the differential equation